Why Utility Maintenance Is Rarely Self-Funding

1
Why Utility Maintenance Is Rarely Self-Funding

• EEI TDM Conference
• Spring 2003
• St. Louis, Mo.

Dan ONeill, Navigant Consulting
2
Executives often ask, How can I spend a dollar
to save a dollar?Or, where is the point when a
dollar of PM saves a dollar of CM?
illustrative
3
The answer is, Dont go there. You wont be
allowed to stayCustomers and regulators will
demand better reliability than that
illustrative
Acceptable reliability
4
It normally costs 10,000 or more to avoid a
tree-caused outageThis figure usually surprises
utility executives
Cost per avoided outage
Outage reduction factor
Outage rate per mile
Trimming cost per mile
2,000 ( .40 x 50
) 10,000 Or, for comparability to capital
projects, take the present value of 2,000 every
four years at a 10 discount rate 5,700
( .40 x 50 )
28,500

• Notes
• Outage rate should be what it would be if not
  trimmed at the normal end of cycle, e.g., fourth
  year, which is higher than the average
  tree-caused outage rate
• Trimming cost is for contact trimming only, not
  removal
• Effectiveness rate recognizes some skips, fast
  growth, limited permissions, and non-contact
  outages like fallen tree/broken limb

5
But, it normally costs only about 1000 to
restore an outageSo, the PM cost is nowhere near
the first-year CM savings
Resource Per Unit Need Per Resource
Resource Units Rate Outage Cost Trouble
responder 1 hour 100 1.0 100 Repair crew (2
x) 4 hours 200 0.4 320 Call taker .1
hour 30 10.0 30 Additional switching 1
hour 100 0.2 20 Additional repair crew 4
hours 200 0.1 80 Sub-total average
labor 550 Material fuse 1 50 1 50 Materia
l pole, X-arm, etc. 1 2000 .2 400 Sub-total
average material 450 Total cost 1000

• Notes
• Time for supervisor, dispatcher, etc., included
  in loaded labor rate
• Although some outages are more severe, often they
  are not very preventable, either

6
Taking credit for all future outages still
doesnt get you thereIt puts the break-even
point within sight, but still just out of reach
Present value of all future avoided CM
Discount rate
Annual avoided cost of outages
Present value of current and future PM
1,000 .10 10,000
28,500
Implies a 10-year payback

• Notes
• The discount rate reflects the weighted average
  cost of capital (plus risk premium, if any)
• Higher costs of capital in the past, e.g., 15,
  were associated with higher rates of inflation
• Dividing by the discount rate is a shorthand way
  of getting the present value of a stream of
  income. It assumes a perpetual annuity. For
  30 years instead, divide by .106 1/9.43
• If you assume 10 includes 3 inflation for the
  annual outage costs, then the discount rate would
  be .07, and the PV would be almost 15,000, but
  not if offset it with a 3 risk premium
• Many executives are not comfortable with a
  10-year payback even though the financial
  analysis says that it is appropriate for a 10
  cost of capital

7
For URD cable replacement, the situation is just
as badEven if the repair cost is assumed to be
twice as high (unless you only replace one
section at a time when its failure rate is bad
enough)
Cost per avoided outage
Outage reduction factor
Outage rate per section
Replacement cost per 200 section
6,000 ( .40 x 100
) 15,000 Or, for entire loops or
subdivisions 6,000 ( .10
x 100 ) 60,000

• Notes
• Outage rate is based on the typical rule of
  replacing a segment if it fails 3 times in the
  last ten years, adjusted to 33 worse for
  increasing deterioration
• Replacement cost is based on 30 per foot on a
  200 segment, replacing direct buried cable with
  direct buried cable, plus lawn restoration.
  Replacing with conduit and cable would be more
  expensive. Injecting with silicon would be about
  half as expensive where it is feasible, but
  usually requires multiple sections, not all of
  which may be failing at a high rate.
• Each time one section of URD cable in a half-loop
  fails, all customers in that half-loop see an
  outage (and all in the other half see one during
  switching), so the outage rate of the loop is
  often much worse than that of any section

8
Likewise for lightning mitigation, etc.And
remediation of other low-incidence pole-top
causes
Cost per avoided outage
Outage reduction factor
Outage rate per pole
Mitigation cost per pole
1,500 ( .05 x 50
) 60,000 1,500 ( .10
x 50 ) 30,000

• Notes
• Outage rate is based on 2 lightning-caused
  outages per mile and 40 poles per mile
• Mitigation cost is based on a three-arrester bank
  and/or improved grounding
• Effectiveness rate is based on reported field
  experience, some of which is even lower, because
  lightning will tend to find the weakest link in
  mitigation
• In order for the PV of the avoided CM cost to
  equal the PV of the PM investment, the outage
  rate per pole would have to be .3, or about every
  3rd pole, or 12 per mile for that cause of outage
  alone, e.g., lightning

9
Part of the problem is that utilities do not
really have a sharp pencilTypical failure
rates on remediation programs are too low
5 failure rate of replaced equipment
Replace 20 units to avoid one failure
If we knew more precisely which unit or mile to
replace or remediate, we could make preventive
maintenance more cost-effective
.2 avoided outages per mile remediated
Remediate 5 miles of line to avoid one outage
5 reject rate on pole inspection
Inspect 20 poles to find one to replace

• Notes
• Outage rates per mile for worst-circuit programs
  sometimes reach high levels, but if that were all
  you did, overall reliability would probably keep
  getting worse
• Replacing failure-prone equipment on a system
  wide basis is rarely cost-effective. You have to
  target the highest failure rate (worst first)
• Information is usually lacking to discriminate
  more finely among classes of equipment to find
  the one that is most likely to fail next

10
One solution is to recognize the business value
of reliability At a value to the utility of
about 25 per customer interruption per year
Reactive Response Cost 2.5 Mil./yr.
PQ Rel. Cust. Sat. 2.8
Customer Interruptions -100,000
Overall Cust. Sat. .64
SAIFI -.14
-100,000 CI / 700,000 Cust. -.14 SAIFI
PQ Rel. Cust. Sat. 23 of Overall Cust. Sat.
Reactive Response 4 Mil./yr. per Cust. Sat. Pt.
10 Points PQ Rel. Cust. Sat. per .5
SAIFI
I.e., a 10-point drop in utilitys customer
satisfaction would require a 40 million response
by utility
Source JD Power Associates, with Navigant
Consulting
11
Many utilities are already seeing such reactive
response costsPenalties/refunds of 25 to 100
per customer interruption are appearing

• Customer refund programs (In addition to
  reliability improvement programs)
• ComEd Commitment 60-100 per customer
  interruption over 8 hours
• IPL refund 100 per customer interrupted over
  24 hours
• PacifiCorp guarantee Up to 100 per customer for
  missed service levels
• Entergy-TX refund 33 per customer per year (for
  120,000 customers)
• PBR Programs (Each of these companies agreed to
  penalties for SAIDI/SAIFI)
• All NY State Utilities (ConEd, National Grid,
  Central Hudson, Energy East, etc.)
• All MA Utilities (NSTAR, National Grid, NU
  (WMECO)
• All CA Utilities (PGE, SCE, SDGE)
• Xcel (in NSP/New Century merger agreement)
• Western Resources (in WR/KCPL proposed merger
  agreement)

12
With such costs included, PM can be
cost-effectiveAnd utilities can go where a
dollar of PM can save a dollar of cost
illustrative
Note At 25 per customer interruption, the value
per outage would be 2,500 - 25,000 per outage,
depending on whether the outages were on
100-customer taps or 1000-customer feeder
backbones. And the PV of all future avoided
outages would be ten times that.
13
Such values can be built into a funding curve
methodologyNavigant Consultings methodology
incorporates this concept
Vertical axis shows cumulative value of projects
to company
Each project is shown adding to totals, ranked by
value/cost ratio

• Option Development
• Developing
• cost-effective alternatives for possible funding
• Additions
• Upgrades
• Replacement
• Maintenance
• Standards
• Systems

• Results Monitoring
• Measuring managing the drivers of the funded
  projects and processes
• Benchmarking
• Unit costs
• Failure rates
• Event impacts
• Value added

Horizontal axis shows cumulative project cost
Exempt
14
Any questions?